Workslop is oozing into every corner of America's white-collar offices (www.businessinsider.com)

Researchers at Stanford Social Media Lab and BetterUp coined the term "workslop" to describe AI-generated deliverables that look polished—slides, reports, emails, or code—but lack real substance. Surveying 1,150 U.S. desk workers, they found 40% reported receiving workslop and that recipients spent nearly two hours on average cleaning up each instance. The phenomenon erodes trust and redistributes effort: instead of fewer hours, many employees now spend more time reviewing, fact‑checking, and reworking AI outputs, which harms team dynamics and can create hidden “AI clean‑up” jobs. The technical implications are stark: generative tools can accelerate output but not reliably improve quality or correctness. Studies cited include Uplevel’s finding that developers using Copilot introduced bugs 41% more often, while GitHub’s internal test showed marginal gains in error‑free lines of code (18 vs. 16). A MIT Media Lab review found 95% of AI pilots yielded no measurable savings. Market behavior reinforces this: Upwork reports rising demand for human editors, designers, and fact‑checkers even as firms expect faster, cheaper work. Bottom line — generative AI can boost throughput but only when humans remain in the loop to provide domain expertise, rigorous review, and context-aware editing; otherwise “workslop” creates more friction than value.